Host Genetic Variations and Sex Differences Potentiate Predisposition, Severity, and Outcomes of Group A Streptococcus-Mediated Necrotizing Soft Tissue Infections

Abstract

Host genetic variations play an important role in several pathogenic diseases, and we previously provided strong evidence that these genetic variations contribute significantly to differences in susceptibility and clinical outcomes of invasive group A Streptococcus (GAS) patients, including sepsis and necrotizing soft tissue infections (NSTIs). The goal of the present study was to investigate how genetic variations and sex differences among four commonly used mouse strains contribute to variation in severity, manifestations, and outcomes of NSTIs. DBA/2J mice were more susceptible to NSTIs than C57BL/6J, BALB/c, and CD-1 mice, as exhibited by significantly greater bacteremia, excessive dissemination to the spleen, and significantly higher mortality. Differences in the sex of the mice also contributed to differences in disease severity and outcomes: DBA/2J female mice were relatively resistant compared to their male counterparts. However, DBA/2J mice exhibited minimal weight loss and developed smaller lesions than did the aforementioned strains. Moreover, at 48 h after infection, compared with C57BL/6J mice, DBA/2J mice had increased bacteremia, excessive dissemination to the spleen, and excessive concentrations of inflammatory cytokines and chemokines. These results indicate that variations in the host genetic context as well as sex play a dominant role in determining the severity of and susceptibility to GAS NSTIs.

INTRODUCTION

Streptococcus pyogenes or group A streptococci (GAS) are the causative agents of a multitude of human diseases ranging from nonsevere strep throat, pharyngitis, and impetigo to life-threatening necrotizing soft tissue infections (NSTIs) and streptococcal toxic shock syndrome (STSS) (1–12). The multifaceted nature of invasive GAS infections requires several modes of pathogenic adaptation to evade host immune defenses to successfully colonize and survive in host niches. In addition to the plethora of pathogenic factors contributing to disease severity, variations in the host genetic context play an equally important role in manipulating disease severity, manifestations, and outcomes.

The emergence of virulent strains of GAS bacteria in the early 1980s coincided with a remarkable resurgence of severe and often deadly forms of invasive infections associated with STSS and NSTIs (5, 6, 12, 13). One particular strain that emerged during that time is the clonal M1T1 strain that disseminated globally and continues to be one of the most prevalently isolated strains from patients with invasive and/or noninvasive GAS infections (12, 14–17). This strain produces many secreted and surface-bound virulence factors; most of them are adapted to evade human host defense mechanisms (12, 18). However, previous studies from our laboratory characterized indistinguishable, invasive M1T1 isolates from patients with starkly different disease severity and outcomes (19). These findings underscored the contribution of host factors to the stark differences in severity and outcomes of invasive GAS infections.

Indeed, ensuing epidemiological studies of large cohorts of infected patients revealed that human leukocyte antigen (HLA) class II allelic variations contribute to important differences in the severity of GAS sepsis by differentially presenting GAS superantigens (SAgs) to T-cell receptor (TCR) Vβ elements and thereby modulating host responses to SAgs that bind simultaneously to TCR Vβ elements and to HLA class II molecules, which are expressed on the surface of host cells. These host responses allow the interactions and exchange of intracellular signals that elicit potent inflammatory responses (20–23). Inability to contain these responses results in severe sepsis, substantial morbidity, and, in many cases, death. Additional studies of humanized HLA class II transgenic mice enabled us to corroborate the role of SAgs in severe invasive GAS infections and revealed that genetic variations in host HLA class II molecules that present SAgs to T cells expressing variable TCR Vβ elements differentially potentiate severity, manifestations, and outcomes of GAS sepsis (24, 25).

Ensuing studies using distinct, conventional, inbred mouse strains corroborated the contributions of variations in host genetic factors in manipulating the severity and outcomes of GAS sepsis (26–28). However, because of the restricted genetic variations in traditional inbred mouse strains, we used genetically diverse panels of recombinant inbred BXD mouse strains whose genetic variations resemble those seen in humans and that will therefore be more likely to yield results that can be translated into clinical practice and thus provide a better understanding of the pathogenesis of sepsis in patients (29, 30). By investigating the contributions of host genetic variation to disease severity and outcomes and by using the robust translational model of genetically diverse BXD mouse strains, we mapped two quantitative trait loci to mouse chromosomes 2 and X, identifying additional host genetic factors associated with the severity of GAS sepsis (29). These studies revealed that in addition to genetic factors, other factors, including age, but not the sex or weight of the infected host, play an important role in modulating the susceptibility to and severity of GAS sepsis (30).

We undertook the present study to investigate whether variation in host genetics would also modulate the severity and manifestations of GAS-mediated NSTIs. To this end, we infected three inbred strains and one outbred strain of mice with the clonal M1T1 strain of GAS and then analyzed inter-mouse strain variation in the severity and outcomes of GAS NSTIs. We report here that, among the strains tested, DBA/2J mice were most susceptible to GAS NSTIs. In addition, mouse sex played a major role in potentiating the susceptibility to and severity of NSTIs. For example, female DBA/2J mice were more resistant to NSTI than were male DBA/2J mice. Furthermore, at 48 h after infection, DBA/2J mice had greater bacteremia, greater bacterial dissemination to the spleen, and increased concentrations of inflammatory cytokines and chemokines. Taken together, these findings underscore the contributions of host genetic context and sex differences in modulating host-mediated responses to GAS NSTIs and thereby potentiating differential susceptibility.

MATERIALS AND METHODS

Ethics statement.

All animal experiments were approved by the Institutional Animal Care and Use Committee (IACUC) at the University of North Dakota.

Bacteria and culture media.

The representative M1T1 clonal GAS isolate 5448WT (19) was routinely grown statically at 37°C in THY medium (Todd-Hewitt broth [Difco] supplemented with 1.5% [wt/vol] yeast extract) as described previously. Sheep blood agar (Becton Dickinson, Franklin Lakes, NJ) was used as solid medium.

Animal studies.

Age- and sex-matched, 8- to 12-week-old mice of three inbred strains (BALB/c, C57BL/6J, and DBA/2J) and one outbred strain (CD-1) were infected subcutaneously with 5448WT; the dose per mouse was 10⁸ CFU in 100 μl of sterile phosphate-buffered saline (PBS). Control animals were injected with sterile PBS alone. After infection, each mouse was housed in an individual cage to avoid contact with or influence from skin lesions of other infected mice. Animals were observed twice a day for a period of 7 days to identify those that died and to look for changes in weight and skin lesions. Animals were humanely euthanized if needed and at specified time points (either 48 h after infection or during the 7-day infection timeline) for cardiac puncture to obtain blood for plasma separation and bacterial load estimations. Necrotic skin tissues and spleen samples were collected and homogenized by a rotor stator homogenizer (Omni International, Marietta, GA) for analysis of bacterial load and dissemination.

Cytokine analyses.

Blood collected from DBA/2J and C57BL/6J mice 48 h after infection with 5448WT was centrifuged at ≥2,000 × g for 3 min to remove the plasma. The concentrations of cytokines and chemokines in these plasma samples were measured by using a Bio-Plex Pro Mouse Cytokine 23-Plex assay kit (Bio-Rad, Hercules, CA) according to the manufacturer's protocol.

Histological analyses.

For histological analyses, at 48 h after infection with 5448WT, the necrotic skin tissues were excised from DBA/2J and C57BL/6J mice and immediately fixed in 10% neutral buffered formalin for 24 h. The samples were then embedded in paraffin blocks, which were processed to obtain 4-μm sections and stained with Mayer's hematoxylin (Sigma-Aldrich, St. Louis, MO) and eosin (Leica Biosystems, Inc., Richmond, IL) staining system according to the manufacturer's protocol. The stained slides were examined using an EVOS FL AutoCell imaging system (Life Technologies, Grand Island, NY).

Statistical analyses.

Statistical analyses were performed by Prism v6.0d (GraphPad, La Jolla, CA). Survival analyses were done with the log-rank (Mantel-Cox) test. One-way analysis of variance (ANOVA) with Bonferroni's post hoc tests was used to evaluate differences in log-transformed bacterial load (skin, blood, and spleen) among the four strains. Log-transformed bacterial load differences between males and females were examined by using unpaired, two-tailed t tests. Two-way ANOVA with Bonferroni's post hoc tests was used to evaluate differences in both percent weight change and change in lesion sizes among the four strains over time. The critical significance value (α) was set at 0.05, and if the P values were less than α, we reported that, by rejecting the null hypothesis, the observed differences were statistically significant.

RESULTS

DBA/2J mice are more susceptible to GAS NSTIs than the other strains tested.

To investigate the ability of host genetics to influence GAS NSTIs, we selected three inbred mouse strains (BALB/c, C57BL/6J, and DBA/2J) and one outbred mouse strain (CD-1) and infected each subcutaneously with equal doses of 5448WT bacteria as described in Materials and Methods. Marked differences in susceptibility (survival) were observed (Fig. 1a). DBA/2J mice were more susceptible to GAS NSTIs than the other strains. The survival rate of DBA/2J mice was 15%, whereas the other three strains—CD-1, BALB/c, and C57BL/6J—had survival rates of 90, 75, and 72.7%, respectively. To identify whether there were differences in bacterial load (in skin and blood) and bacterial dissemination between the mouse strains, we analyzed blood, skin, and spleen samples from surviving and dead mice during the 7-day infection timeline as described in Materials and Methods. DBA/2J mice had greater bacterial loads in the skin and blood and greater bacterial dissemination to the spleen than did the three other strains of mice (Fig. 2a).

FIG 1.

Differential susceptibility to GAS due to host genetics and sex differences. Mice were infected subcutaneously with 10⁸ CFU of GAS isolate 5448WT. Survival curves for the four strains C57BL/6J, DBA/2J, BALB/c, and CD-1 (a) and for C57BL/6J (b) and DBA/2J (c) males and females are shown. The data presented are the percent survival (n ≥ 20 for each strain or n ≥ 10 for each sex from a total of three experiments). P values were calculated by log-rank (Mantel-Cox) tests. NS, not significant; *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

FIG 2.

Differential bacterial loads in response to GAS NSTIs. To enumerate bacterial loads and dissemination, we collected skin tissues, blood, and spleens from surviving and dead mice of strains C57BL/6J, DBA/2J, BALB/c, and CD-1 (a) and from strain C57BL/6J (b) and DBA/2J (c) males and females during the 7-day infection timeline as described in Materials and Methods. Each symbol represents one mouse, each horizontal bar denotes the mean value for each strain or sex, and the horizontal dotted line indicates the inoculum given. For dead mice for which we did not collect blood (×), the bacteremia counts were given an arbitrary value of 10¹⁰ CFU/ml (the value closest to the maximum bacteremia). The data presented are log-transformed bacterial loads (n ≥ 20 for each strain or n ≥ 10 for each sex from a total of three experiments). P values were calculated using one-way ANOVA (among four strains) with Bonferroni's post hoc analyses (among C57BL/6J and DBA/2J mice) and unpaired, two-tailed t tests (between males and females of the indicated strains). NS, not significant; *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Resistant mice recover faster.

We next analyzed whether marked differences were seen in any additional comorbidities (i.e., weight loss and lesion area) between the four strains. Mice began to die starting day 3 after infection. Mice that died had consistently lost weight, whereas resistant strains started gaining weight beyond 4 days after infection. Much to our surprise, compared to the other mouse strains, the susceptible DBA/2J mice lost minimal weight through day 4 after infection. Nevertheless, there was a significant difference in weight loss between DBA/2J and the rest of the mouse strains, in particular C57BL/6J (P < 0.0001), from days 2 to 4 after infection (Fig. 3a). Beyond day 4, the weight loss differences were not significant between DBA/2J and the other strains, and the lack of significant differences at this time may be due to the weight gain in the resistant strains but not in DBA/2J mice (Fig. 3a). In contrast, lesion size differed significantly between DBA/2J and C57BL/6J mice only on day 3 after infection (Fig. 3d). Similar to the weight loss results, the NSTI lesions of susceptible DBA/2J mice were minimal in size compared to those of the other strains until day 4 after infection. However, the size of the lesions in DBA/2J mice continually increased past day 4 after infection, whereas the size of lesions in resistant strains reached a plateau. In general, these observations revealed that the susceptible mice continually lost weight and developed lesions of increasing size, whereas the resistant mice tended to recover more quickly.

FIG 3.

Resistant groups heal faster. Shown are the percent weight changes and lesion sizes over the 7-day infection timeline for the four strains (a and d) and for C57BL/6J (b and e) and DBA/2J (c and f) male and female mice. Error bars indicate means ± the standard errors of the means (SEM) (n ≥ 20 for each strain or n ≥ 10 for each sex from a total of three experiments). P values were calculated by two-way ANOVA, and if an interaction (P < 0.05) was revealed, then Bonferroni's post hoc analyses were computed at different time points between C57BL/6J and DBA/2J (a and d) or between male and female mice of indicated strains (c and f). NS, not significant; *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Female DBA/2J mice are resistant to GAS NSTIs.

To assess whether sex has a role in mouse susceptibility to GAS NSTIs, the survival rates of male and female DBA/2J mice were compared. Among the susceptible DBA/2J mice, the percentage of surviving females was significantly greater than the percentage of surviving males of the same strain (Fig. 1c). Further, we investigated the extent of bacterial load and dissemination within DBA/2J female mice. They had lower bacterial loads along with less dissemination to the spleen as shown in Fig. 2c. Weight loss between females and males differed significantly, with females losing minimal weight past day 3 after infection (Fig. 3c). Similarly, DBA/2J female mice had relatively smaller lesions than did male mice (Fig. 3f). In contrast, within one of the resistant strains (C57BL/6J), significant differences between the male and female mice were not observed (Fig. 1b, 2b, 3b, and 3e).

DBA/2J mice exhibit ineffective GAS clearance from the blood at 48 h after infection.

We hypothesized that ineffective bacterial clearance was responsible for the observed susceptibility to NSTIs in DBA/2J mice. To test this hypothesis, we injected equal doses of GAS subcutaneously into DBA/2J mice and C57BL/6J mice (one of the resistant strains). At 48 h after infection, mice were bled, and samples of their infected skin and spleens were excised for analysis of bacterial load and dissemination. Compared to C57BL/6J mice, DBA/2J mice had no significantly different bacterial load at the site of infection (skin), but the bacterial load in blood and dissemination to spleen in DBA/2J mice were significantly greater than those in C57BL/6J mice (Fig. 4a). Nevertheless, the female DBA/2J mice had no bacteria in the blood at 48 h after infection (Fig. 4b). These observations indicate that the continual persistence of bacteremia due to ineffective bacterial clearance might account for the increased dissemination and subsequent death of DBA/2J mice. Despite the increased bacteremia and dissemination, further analyses of weight loss and lesion area revealed that DBA/2J mice had a significantly lower percentage of weight loss and smaller lesions than did C57BL/6J mice at 48 h after infection (Fig. 4b and c). Also, we did not observe any significant differences in weight loss or lesion size between male and female DBA/2J mice at 48 h after infection (see Fig. S1 in the supplemental material). These observations further corroborate the results shown in Fig. 3.

FIG 4.

DBA/2J mice do not efficiently clear bacteremia but still display minimal weight loss and smaller lesions than do C57BL/6J mice. At 48 h after infection with ∼10⁸ CFU of GAS isolate 5448WT, skin tissue, blood, and spleen samples were collected and homogenized to enumerate bacterial loads and dissemination to the spleen in C57BL/6J and DBA/2J mice (a) and bacteremia levels in male and female mice of the DBA/2J strain (b). Each symbol represents one mouse, each horizontal bar denotes the mean value for each strain or sex, and the horizontal dotted line indicates the inoculum given. The data presented are log-transformed bacterial loads (n = 19 for each strain or n ≥ 9 for each sex from a total of three experiments). (c) The percent weight changes and lesion sizes in C57BL/6J and DBA/2J mice are shown. Error bars indicate means ± the SEM (n = 19 in each strain from a total of three experiments). P values were calculated by unpaired, two-tailed t tests for bacterial enumerations (a and b) and two-way ANOVA with Bonferroni's post hoc analyses for weight loss and lesions (c). NS, not significant; *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

DBA/2J mice produce greater concentrations of inflammatory cytokines and chemokines in response to GAS NSTIs than C57BL/6J mice.

We previously described host-dependent differences in the levels of cytokine responses during severe invasive GAS infections (21, 22, 24). To test whether the same scenario was observed in DBA/2J versus C57BL/6J mice, we analyzed the plasma recovered from these mice at 48 h after infection by using multiplex cytokine and chemokine analyses. As shown in Fig. 5, there was a significant increase in concentration in both inflammatory cytokines (interleukin-6 [IL-6] and IL-12) and inflammatory chemokines (CXCL1/KC and CCL2/MCP-1) in DBA/2J mice. However, the concentration of tumor necrosis factor alpha (TNF-α) was slightly higher in C57BL/6J mice (Fig. 5e). However, we did not see any significant differences in cytokine concentrations between male and female DBA/2J mice at 48 after infection (Fig. 6).

FIG 5.

DBA/2J mice produce greater concentrations of inflammatory cytokines and chemokines than C57BL/6J mice. Shown are the results of the differential analyses of IL-6 (a), IL-12 (b), keratinocyte-derived cytokine (KC) (c), monocyte chemoattractant protein 1 (MCP-1) (d), and TNF-α (e) between GAS-infected C57BL/6J and DBA/2J mice at 48 h after infection. Error bars indicate means ± the SD (n = 19 for each strain from a total of three experiments). P values were calculated by unpaired, two-tailed t tests. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

FIG 6.

No significant differences in cytokine and chemokine production between males and females of DBA/2J strain in response to GAS NSTIs. Shown are the measured cytokine levels of IL-6 (a), IL-12 (b), KC (c), MCP-1 (d), and TNF-α (e) between GAS-infected male and female mice of DBA/2J strain at 48 h after infection. Error bars indicate the means ± the SD (n ≥ 9 for each sex from a total of three experiments). The P values were calculated by unpaired, two-tailed t tests. NS, not significant.

Female mice exhibit more adipose cells in response to GAS NSTIs than male mice.

A recent study has demonstrated that in response to Staphylococcus aureus skin infections in C57BL/6J mice, dermal preadipocytes proliferate and protect against the infection (31). In addition, a previous study has shown that sex plays an important role in skin morphology and physiology (32). Based on these studies, we hypothesized a similar proliferation against GAS skin infections; also, we wanted to determine whether there are any marked sex-based differences in this proliferation. To investigate this, necrotic skin tissues excised from male and female mice of DBA/2J and C57BL/6J strains at 48 h after infection with GAS bacteria were stained using hematoxylin and eosin. We observed that in response to GAS infection, female mice exhibited more adipocytes than their male counterparts (Fig. 7).

FIG 7.

Comparison of GAS-infected skin tissues from male and female mice of DBA/2J and C57BL/6J strains. At 48 h after infection with ∼10⁸ CFU of GAS isolate 5448WT, necrotic skin tissues excised from males and females of DBA/2J and C57BL/6J strains were stained with hematoxylin and eosin. Shown are representative pictures of stained skin samples from DBA/2J (a to d) and C57BL/6J (e to h) male and female mice. Epidermis (E), dermis (D), and panniculus carnosus (PC) are indicated. Scale bar, 200 μm.

DISCUSSION

In this study, we demonstrate the novel finding that host genetic variations and sex differences predispose mice to GAS-mediated NSTIs. DBA/2J mice were relatively susceptible to GAS NSTIs: we observed increased mortality, ineffective bacterial clearance, increased bacterial loads and dissemination, elevated inflammatory markers, and poor recovery from the comorbidities of weight loss and lesions. Further, female DBA/2J mice were more resistant to GAS NSTIs than were male mice of the same strain. This finding underscores the necessity of incorporating differences in the host genetic context and sex in studies investigating factors affecting host-pathogen interactions in GAS NSTIs.

Longitudinal and endpoint analyses of the comorbidities of weight loss and lesion size revealed that these two are not reliable markers for disease progression. For instance, the most susceptible DBA/2J mice lost minimal weight and developed smaller lesions during the initial stages of infection, despite exhibiting greater bacterial loads and dissemination and increased concentrations of inflammatory cytokines and chemokines. Nevertheless, continued loss of body weight and an increase in lesion size were observed in susceptible mice throughout the infection timeline.

Several investigators have successfully reproduced GAS NSTIs in mice by utilizing conventional inbred strains, including BALB/c (33, 34) and C57BL/6J (35, 36) mice, and outbred strains, such as CD-1 (37–39) and hairless Crl:SKH1-hrBR (40, 41) mice. However, most of these studies utilized female mice alone; therefore, the underlying role of sex differences could not be examined in relation to host genetic variations in differential susceptibility to GAS NSTIs. To our knowledge, none of the reported studies of GAS NSTIs have utilized DBA/2J mice, although these mice have been used in the studies of GAS sepsis (26–30).

The role of host genetics and sex differences in GAS NSTIs has not been reported so far; however, the role of host genetics but not sex in GAS sepsis has been widely reported (22, 23, 26–30). For example, we previously documented that human leukocyte antigen (HLA) class II variations can contribute to differences in host responses to GAS SAgs (22, 23). Further, our application of a systems genetics approach to identify additional host factors that modulate GAS sepsis revealed that the IL-1α and prostaglandin E synthase pathways are involved in this complicated pathogenesis (29, 30). Moreover, host contributions to pathogenesis of other infectious agents, including Ebola virus (42), herpes simplex virus type 1 (43), influenza H1N1 virus (44), and Burkholderia pseudomallei (45), have also been widely studied. Taken together, we can conclude that host-mediated pathogenesis has been a topic of thorough investigation; however, to our knowledge, the contribution of host genetic variations and sex differences to responses to GAS NSTIs in particular is a novel observation that we report here for the first time.

In summary, the results presented here provide evidence for the role of host genetics and sex in susceptibility to GAS NSTIs. Further in-depth systems genetics-based investigations, which are currently ongoing in our laboratory with the utilization of the advanced recombinant inbred BXD panel, will provide a thorough understanding of the mechanism(s) by which host genetics and other nongenetic cofactors (including sex, age, and body weight) may affect host-pathogen interactions in the context of GAS NSTIs and should expose means by which to prevent and/or ameliorate excessive morbidity and mortality associated with these infections.

Funding Statement

A startup grant from the University of North Dakota provided funding to Malak Kotb. The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.